A significant share of the population in the Kyrgyz Republic – 37 percent – lived below the poverty line in 2011, according to the latest available data. And despite a relatively modest population of about 5.5 million, poverty rates across oblasts (provinces) span a striking range -- from 18 percent to 50 percent.
Why? Well, that is a surprisingly difficult question to answer.
Could it be that people with favorable characteristics such as more education, full-time employment, and fewer dependents are concentrated in specific areas of the country? To some extent, we find that households with “better” characteristics do indeed live in Bishkek compared to other urban and rural areas. But even then, only 54 to 64 percent of the variation in welfare (once we control for characteristics) can be explained by these factors.
But why do people with similar characteristics have different incomes? It could be that there are some unobservable household characteristics, which the analysis is unable to discern – like the quality of education. Another explanation could be that there is something about geographical location that affects people’s productivity—and consequently, their incomes.
To better understand this variation, we created a “poverty map” of the Kyrgyz Republic that estimates poverty rates at the rayon (district) level. This not only paints a clearer picture of how poverty looks across the country, but allows us to explore the impact of location-specific factors on the likelihood of being poor.
One extension of this work is to determine the importance of “market accessibility” of the population. Rayons with high market accessibility are close to large population centers and have access to good quality roads. We find that such rayons have lower poverty rates, thus confirming our assumption that the transport infrastructure matters for welfare. Though more difficult to quantify, accessibility to the transport network could also improve welfare by giving households better access to schools, health clinics, and other essential public services.
Finally, the poverty map opens up several possibilities to try and understand why some areas are better off than others. The World Bank Group has begun exploring this topic. We hope to understand to what extent the amount and type of government expenditures effect people’s welfare at the rayon level. Other avenues to explore are whether wealthier rayons have access to areas with more business activity.